AutodE unable to find NWChem about autode

[eneas77]

Finally this was sorted including full path of the executable mpirun in NWChem.py
For some reason anaconda mpirun one it wasn´t working properly...

Thank you!!

[t-young31]

Hi! Thanks for asking 😄

It looks like autodE has been able to find the NWChem install, its just been unable to get the optimised set of coordinates (atoms) from the output file. What does the H2_opt_nwchem.out file look like?

[eneas77]

Hi!

No protocol specified
No protocol specified
No protocol specified
No protocol specified

That is because I have a built version of nwchem that works with openmpi... probably it is a matter of configure and detect mpirun?

Cheers,

[t-young31]

That will be it. I didn't realise NWChem could be built like that, which means unfortunately it's only called with mpirun (see this line). If you've got access to an MPI version or configure it to use mpirun then that's probably the easiest solution, otherwise how do you invoke the openmpi version? and I'll update the invocation to support that 👌

[t-young31]

If this is resolved then please feel free to close this issue 😄